油菜茎基溃疡病菌在中国定殖的可能性评估
Assessment of potential establishment of fungus Leptosphaeria maculans in China

为明确油菜茎基溃疡病菌Leptosphaeria maculans在我国定殖的可能性,并制订针对性的检疫措施,本研究基于该病菌在英国、法国、德国、波兰、加拿大、澳大利亚的分布数据,选取与病害发生有关的温度、降水量等15个变量,利用MaxEnt和GARP两种生态位模型预测其在其它地区的潜在分布,并利用受试者工作特征曲线(ROC)来评价模型模拟精度。结果显示,2种模型均能够较好地预测L. maculans的分布区域,与已记载的分布区域高度吻合,且全球范围内存在着许多高度适合L. maculans定殖的地区;2种模型也能预测L. maculans在我国的潜在分布区域,并且预测结果一致;GARP结果显示L. maculans在我国的中高度适生区包括内蒙古、吉林、陕西、宁夏、甘肃、新疆、西藏等地。鉴于L.maculans在我国的潜在分布研究结果,建议完善疫情监测体系,采取措施控制病菌的"进入风险",降低病菌"进入"后定殖的可能性。
To define the potential establishment of Leptosphaeria maculans, the causal agent of blackleg disease of brassicas, in China and implement phytosanitary measures, fifteen eco-geographical variables including temperature and precipitation were selected for the ecological niche model analysis. Maximum Entropy Method (MaxEnt) and Genetic Algorithm for Rule-set Prediction (GARP) were used to simulate the potential distribution areas of L. maculans based on the sampling distribution records in England, French, Germany, Poland, Canada and Australia. The simulation precision of the models was assessed by the method of Receiver Operating Characteristic (ROC) curve analysis. The results showed that both MaxEnt and GARP models could predict the suitable distribution areas of L. maculans, which were highly matched to the occurrence records. And there were plenty of highly suitable regions for colonization of L. maculans in the globe. The two models also could forecast the potential distribution regions in China, and the predictions were coincident with each other. The result of GARP indicated that the areas favorable for the establishment of L. maculans in China were Inner Mongolia, Jilin, Shaanxi, Ningxia, Gansu, Xinjiang and Tibet. Therefore, corresponding quarantine and surveillance measures should be taken to control the risks of the introduction and establishment of L. maculans in China.